Hi,
On Mon, 2 Jul 2012, Mike Stump wrote:
> On May 26, 2012, at 8:03 PM, Michael Matz wrote:
> > -----------------------
> > PR middle-end/38474
> > * cfgexpand.c (struct stack_var): Add slot_type member.
> > (add_stack_var): Initialize it.
> > (add_alias_set_conflicts): Remove.
> > (merge_stack_vars_p, more_specific_type): New functions.
> > (nonconflicting_type): New static data.
> > (union_stack_vars): Use more_specific_type to update slot_type.
> > (partition_stack_vars): Call merge_stack_vars_p ...
>
> So this patch seems to have killed
> 'RUNTESTFLAGS=dg.exp=class_array_3.f03' check-fortran for me. Any
> insight?
Hmm, no. Works for me just fine :-/ So something is shared on the stack,
but then reordered so that both entities are live. If you find which file
is miscompiled you can look at the -fdump-rtl-expand-details dumps for
entries like:
Partition 1: size 32 align 16
dest
Partition 0: size 32 align 16
src
If something is shared there will be multiple decls mentioned in the
lines. (Ignore the lines mentioning Partition without a ':', that the
out-of-ssa partitions, not the stack slots).
> Unfortunately the bug I think manifests as generated code in
> the fortran runtime libraries, so I need to go track down the code.
>
> Also, git-svn-id: svn+ssh://gcc.gnu.org/svn/gcc/trunk@188667 seems to
> not match this patch, nor any other patch I can find, but it is the
> closest match.
The patch that went in is
http://gcc.gnu.org/ml/gcc-patches/2012-06/msg00367.html
Ciao,
Michael.
